本稿ではBlog 情報を先行指標とした書籍需要予測システムを作る事により, 書籍の適正需要数を把握するシステムの構築を行う事を目的とする. そのため, 書籍売上数と書籍名を利用したBlog 検索ヒット数との相互相関分析を行い, 両者の間には何らかの相関がある事を確認した. また, 本によっては先行指標になる可能性があるものの, 書籍名だけでは不十分なものもある事が分かった.

The Input-Output Table summarize dealings between the industries of one country for one year in the table, and it has been used in economic analysis for many years. The subject about mathematical analysis of this table is introduced.

English Abstract In nuclear power plants and chemical plants safety enhancement of systems using research results from safety engineering is being undertaken. However, collection and analysis of information is mostly conducted manually, and further efficiency is desired. Here, plans to use text mining techniques for preparation of safety assessment criteria have been proposed.

Human activity data, so-called "life-log data", has underlying contexts such as sleeping, driving a car, eating. In order to analyse and predict such data, the hidden Markov model is often used. However, the duration time of the hidden context state of HMM distributes exponentially and this is not suited for modeling the above contexts. In order for modeling these context flexibly, we introduce and compare probabilistic modeling techniques of of hidden states with general duration distributions.

So far, many of the computational models for rendering music performance have been proposed. This review gives an overview of computational modeling of expressive music performance, especially using machine learning method.

This survey focuses on tensor factorization methods based on stochastic models. They are natural extensions of the conventional methods to Bayesian statistics and are easier to introduce prior knowledge. Some of their applications are also given.

Ordered lists of objects are widely used as representational forms. Such ordered objects include Web search results or bestseller lists. We survey distances and probabilistic distributions that are designed for orders or rankings.